1. Field of the Invention
The present invention relates to a voltage regulator for providing a regulated supply voltage to a load from an input voltage.
An example of application of the present invention concerns integrated circuits for remote-supply telephone sets where the supply is provided by the telephone line, either by the ring circuit when the set is not picked up, or by the speech circuit when the set is picked up, or even by a supply specific to the telephone set (for example, a battery).
The present invention more specifically applies to a regulator for automatically selecting a highest input voltage from among several uncorrelated supply voltages, that is, voltages issued by distinct circuits and provided on several independent inputs of the regulator.
2. Discussion of the Related Art
FIG. 1 shows a conventional diagram of a regulator for supplying a voltage regulated at a specific value from a single supply voltage.
Such a regulator receives, on an input terminal E, a supply voltage to be regulated V, and provides, on an output terminal S, a regulated voltage V.sub.R. The regulator includes a circuit 15 supplying a reference voltage, and a control circuit 16 for controlling a P-channel MOS power transistor M10, the source of which is connected to terminal E and the drain of which constitutes terminal S. Circuit 15 has the finction of determining a precise reference voltage V.sub.BG for controlling, via control circuit 16, output voltage V.sub.R. Circuit 15 includes two PNP-type bipolar transistors Q1 and Q2, the respective emitters of which are connected to terminal E and the respective collectors of which constitute two output terminals 3, 4, of circuit 15 for controlling circuit 16, as it will be seen hereafter. The bases of transistors Q1 and Q2 are connected to the collector of transistor Q1. The collectors of transistors Q1 and Q2 are respectively connected to the collectors of NPN-type bipolar transistors Q3 and Q4, the bases of which are interconnected and form a terminal 5 at reference potential V.sub.BG. The emitter of transistor Q4 is connected to the ground via two resistors R1 and R2 mounted in series. The emitter of transistor Q3 is connected to the midpoint of the series association of resistors R1 and R2. Resistances R1 and R2 and the surface ratio of transistors Q3 and Q4 are chosen to obtain the desired voltage V.sub.BG with a given current in transistors Q1, Q2, Q3, and Q4. Circuit 15 includes a starting circuit formed of a current source I, the output of which is connected to the ground via a diode D and to the base of an NPN-type bipolar transistor Q.sub.D, the collector of which is connected to terminal 4 and the emitter of which is connected to the midpoint of the series association of resistors R1 and R2.
Circuit 15 shown in FIG. 1 is generally referred to as a "band gap" (BG) circuit and its operation is perfectly well known.
Circuit 16 for controlling transistor M10 is formed of two PNP-type bipolar transistors Q5 and Q6, the respective emitters of which are connected to terminal E and the bases of which are respectively connected to terminals 4 and 3. The collectors of transistors Q5 and Q6 are connected to the respective drains of two N-channel MOS transistors M11 and M3 mounted as a current mirror, the sources of transistors M11 and M3 being connected to the ground and transistor M11 being diode-connected. The collector of transistor Q6 constitutes an output terminal of circuit 16 connected to the gate of transistor M10. A resistive bridge formed by resistors R3 and R4 is generally connected between terminal S and the ground when the desired voltage V.sub.R is different from reference voltage V.sub.BG. The midpoint of their dividing bridge is connected to terminal 5 of circuit 15 to constitute a reverse feedback loop enabling maintenance of reference voltage V.sub.BG on the bases of transistors Q3 and Q4. This reference voltage ensures that the currents in transistors Q3 and Q4 are equal. When there is a drift with respect to this reference voltage, the currents in transistors Q1 and Q2 are unbalanced. This current unbalance is amplified by circuit 16 and modifies potential V.sub.G of control of transistor M10 to reestablish, via resistive bridge R3-R4, voltage V.sub.BG which makes the current in transistors Q3 and Q4 equal. Voltage V.sub.R is equal to V.sub.BG (R3+R4)/R4.
A capacitor C is generally provided at the output of the regulator and is connected between terminal S and the ground. The function of this capacitor is, in particular, to ensure the stability of the reverse feedback loop.
A disadvantage of a regulator such as shown in FIG. 1 is that, if voltage V becomes lower than regulated voltage V.sub.R, terminals E and S are short-circuited by transistor M10. Indeed, the substrate of MOS transistor M10 or its well generally is connected to its source, that is, to potential V. The substrate of a MOS transistor or its well is generally referred to as the "bulk" of the transistor to distinguish it from the general substrate of the integrated circuit whereon are implemented the different components. The bulk of a MOS transistor is generally symbolized by an arrow, the direction of which indicates the P or N type of the transistor channel. When voltage V.sub.R is higher than voltage V, the PN junction between the drain and the bulk of transistor M10 is forward biased and the transistor then is short-circuited by the drain/bulk diode. Further, the drain and the source of transistor M10 exchange (the current being reversed), which turns the reverse feedback operated by circuit 15 into a feedback.
This short-circuiting is prejudicial to a second function of capacitor C, which is to temporarily supply the load in case of a deficiency or a disappearing of supply voltage V. For example, when the regulator is used to supply a microprocessor, it is desired to maintain the supply of the microprocessor for the time required for it to store the data, after a deficiency or a disappearing of the supply voltage. Voltage V.sub.R is generally compared with a threshold by means of a circuit external to the regulator to detect a decrease in voltage V.sub.R and then use capacitor C to temporarily supply the microprocessor before voltage V.sub.R disappears.
A conventional solution to insulate terminal E from the rest of the regulator, when the supply voltage becomes lower than voltage V.sub.R, is to place a diode at the input of the regulator. However, a disadvantage of such a solution is that it introduces a voltage drop of about 0.7 volt between the input and output terminals of the regulator.
Insulating diodes are also used when it is desired to supply the regulator such as shown in FIG. 1 from different voltages by selecting, as the voltage to be regulated, that having the highest potential.
FIG. 2 shows a conventional example of a voltage regulator automatically selecting, among two supply voltages V.sub.M and V.sub.L arriving on two input terminals E.sub.M and E.sub.L, the highest voltage. Circuits 15 and 16 shown in FIG. 1 have been functionally schematized in FIG. 2 by a reference voltage generator 1 and by an amplifier 2 receiving, as an input, reference voltage V.sub.BG and the potential of the midpoint of resistive dividing bridge R3-R4. Amplifier 2 and generator 1 are biased by the highest supply voltage V.sub.M or V.sub.L by means of diodes, respectively D1, D2, and D3, D4 interposed in series between each terminal E.sub.M or E.sub.L and the biasing terminal of generator 1 or of amplifier 2.
If such a circuit does enable selection of the highest supply voltage, the use of diodes has, as previously, the disadvantage of introducing a voltage drop of about 0.7 volt in series with the regulator.
Another solution known in the art consists in using, instead of diodes, MOS transistors suitably controlled and that have the same function of selecting the highest voltage and the function of isolating the lowest voltage. These transistors, like diodes, introduce an additional voltage drop.
European Patent application 0465933 discloses us a voltage regulator adapted to be supplied from a plurality of mutually independent voltages. An amplifier of a voltage proportional to an error voltage between the regulated voltage and a reference voltage controls a multi-emitter bipolar transistor, each emitter of which is connected to a supply voltage. A first drawback of this regulator is that it has high power consumption when the highest of the supply voltages is lower than the desired regulated output voltage. Indeed, the amplifier then tries to maintain the output voltage at the desired value and a high current is drawn through the bipolar transistor. Another drawback is that the supply terminals associated with the lowest voltages are not isolated from the remaining parts of the circuit if they are at least 0.7 volts highest than the control voltage of the multi-emitter transistor. Additionally, in this case, a plurality of supply terminals can be short circuited. Additionally, the use of diodes for supplying the control amplifier causes a significant residual voltage (that is, a minimum voltage between the highest supply voltage and the output voltage), even if the bipolar transistor is replaced by field effect transistors.